Hi,
I can currently not verify this, but anyways: I think that union's were and
possibly are implicitly treated as volatile by SDCC and/or the PIC
backend(s) due to problems in maintaining correct potential alias sets.
This might explain the bad code that is generated for the a+=b operation
you see, as your code does use unions. To test this, you could x+=y two
global uint16_t variables and see what happens. For structs, I would expect
the same code as for globals. It might also be interesting to see what
happens if the right operand is simple (say, just a global variable or even
a function argument):
uint16_t x, y;
struct {
uint16_t v;
uint16_t w;
} s;
union {
uint16_t a;
uint16_t b;
} u;
void foo()
{
x += y;
s.v += s.w;
u.a += u.b;
}
void bar()
{
x += y;
s.v += y;
u.a += y;
}
By the way, I can imagine that += on bitfields is even worse ...
Best regards,
Raphael
On Tue, Jun 24, 2014 at 8:56 PM, Kustaa Nyholm <kustaa.nyh...@planmeca.com>
wrote:
> Right,
>
> sorry about that.
>
> I'm working on PIC18F4550.
>
> Compilable function ... well I can give you the defs and a code
> snippet:
>
> typedef struct {
> union {
> uint16_t nco;
> struct {
> unsigned nco_low_bits :15;
> unsigned nco_hi_bit :1;
> };
> };
> struct {
> unsigned has_next :1;
> unsigned reserve_b1;
> unsigned reserve_b2 :1;
> unsigned reserve_b3 :1;
> unsigned reserve_b4 :1;
> unsigned reserve_b5 :1;
> unsigned reserve_b6 :1;
> unsigned reserve_b7 :1;
> };
> uint16_t speed;
> uint8_t next_steps;
> uint16_t next_speed;
> uint8_t next_dir;
> uint8_t steps;
> } motor_t;
>
>
> #define MOTOR g.motor[0]
>
>
> and the code snippet:
>
> ....
> MOTOR.nco += MOTOR.speed;
> if (MOTOR.steps) {
> if (MOTOR.nco_hi_bit) { // speed NCO overflowed and we have steps left =>
> generate pulse
> MOTOR.steps--;
> ....
>
>
> so basically it is just advancing a 16 bit unsigned variable
> by adding an other 16 bit unsigned to it. I treat the value
> as 15 bit and check the highest bit for overflow, in assembler
> I would have access to the carry bit and could use the full
> 16 bits which would better but I'm a happy to use just 15 bits
> if I can stay in the C-domain.
>
> Sorry that I did not provide this info in the first place,
> I was so deeply into my code that I took all this granted,
> which it isn't of course.
>
> br Kusti
>
>
>
>
>
>
>
> On 24/06/2014 20:40, "Philipp Klaus Krause" <p...@spth.de> wrote:
>
> >-----BEGIN PGP SIGNED MESSAGE-----
> >Hash: SHA1
> >
> >I think it would be helpful to readers of your post, if you'd
> >
> >* State which target this is about
> >* State which data types the variables are (or even better: give a
> >compileable function)
> >
> >While these could be inferred from the assembly code you posted, they
> >would still make reading easier.
> >
> >Philipp
> >
> >-----BEGIN PGP SIGNATURE-----
> >Version: GnuPG v1
> >Comment: Using GnuPG with Icedove - http://www.enigmail.net/
> >
> >iEYEARECAAYFAlOpuBkACgkQbtUV+xsoLprMJQCgnSxz07CLU7dkx1eGdQMK6IGh
> >TBcAnj62xyTYqBXFYQ8qm8nKpyKKshKr
> >=AENS
> >-----END PGP SIGNATURE-----
> >
> >--------------------------------------------------------------------------
> >----
> >Open source business process management suite built on Java and Eclipse
> >Turn processes into business applications with Bonita BPM Community
> >Edition
> >Quickly connect people, data, and systems into organized workflows
> >Winner of BOSSIE, CODIE, OW2 and Gartner awards
> >http://p.sf.net/sfu/Bonitasoft
> >_______________________________________________
> >Sdcc-user mailing list
> >Sdcc-user@lists.sourceforge.net
> >https://lists.sourceforge.net/lists/listinfo/sdcc-user
> >
>
>
> --
> Kustaa Nyholm
> Research Manager, Software
> Research and Technology Division
> PLANMECA OY
> Asentajankatu 6
> 00880 HELSINKI
> FINLAND
>
> Please note our new telephone and fax numbers!
> Tel: +358 20 7795 572 (direct)
> Fax: +358 20 7795 676
> GSM: +358 40 580 5193
> e-mail: kustaa.nyh...@planmeca.com
>
>
>
>
>
>
> This e-mail may contain confidential or privileged information. If you are
> not the intended recipient (or have received this e-mail in error) please
> notify the sender immediately and destroy this e-mail. Any unauthorized
> copying, disclosure or distribution of the material in this e-mail is
> strictly forbidden. We will not be liable for direct, indirect, special or
> consequential damages arising from alteration of the contents of this
> message by a third party or as a result of any virus being passed on or as
> of transmission of this e-mail in general.
>
>
> ------------------------------------------------------------------------------
> Open source business process management suite built on Java and Eclipse
> Turn processes into business applications with Bonita BPM Community Edition
> Quickly connect people, data, and systems into organized workflows
> Winner of BOSSIE, CODIE, OW2 and Gartner awards
> http://p.sf.net/sfu/Bonitasoft
> _______________________________________________
> Sdcc-user mailing list
> Sdcc-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/sdcc-user
>
------------------------------------------------------------------------------
Open source business process management suite built on Java and Eclipse
Turn processes into business applications with Bonita BPM Community Edition
Quickly connect people, data, and systems into organized workflows
Winner of BOSSIE, CODIE, OW2 and Gartner awards
http://p.sf.net/sfu/Bonitasoft
_______________________________________________
Sdcc-user mailing list
Sdcc-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sdcc-user
